With the rise of micro-blog, increasingly more network users become micro-blog users, and there are even a group of users having a long relation chain. A user having a long relation chain refers to a user whose follower number exceeds a certain number, such as a star authenticated user. Because the user having a long relation chain has too many followers and the volume of generated data is huge, how a server improves a speed of responding to a huge number of requests and completes a service as soon as possible is a basic problem that needs to be solved.
In the existing technology, a data buffering layer receives an operation request of a follower of a user, having a long relation chain, of a client, where the operation request includes a follower adding request, a follower modifying request, a follower deleting request, or the like; then, the data buffering layer synchronizes the received operation request to a stream module; after receiving the operation request sent by the data buffering layer, the stream module stores in hours each operation request to a database corresponding to the operation request, and the database completely stores entire follower information of one user having a long relation chain; then, the stream module actively pushes the operation request in each database to a friend module at regular intervals; and after receiving the operation request pushed by the stream module, the friend module performs an operation on follower information in a database corresponding to the user having a long relation chain according to the operation request.
In a process of implementing the present disclosure, the inventor finds that the existing technology at least has the following problems:
Because one database not only stores an operation request, but also stores entire follower information of a user having a long relation chain, data operation actions of the stream module and the friend module are too slow; as a result, a processing speed of the data buffering layer does not synchronize with processing speeds of the stream module and the friend module, thereby resulting in a low speed of processing the follower information of the user having a long relation chain.